An Observational Study Investigating the Need for Decompressive Hemicraniectomy after Thrombectomy in Acute Ischemic Stroke of the Middle Cerebral Artery Territory

Objective: The frequency incidence of decompressive hemicraniectomy following intra-arterial thrombectomy (IAT) in acute ischemic stroke (AIS) involving the middle cerebral artery (MCA) territory was assessed as a surrogate for morbidity. Methods: A single-institution retrospective chart review was conducted involving 209 consecutive patients between September 2014 and May 2017 with infarctions affecting the MCA territory and who subsequently underwent IAT. The outcomes of interest included the frequency of hemicraniectomy following IAT and the effects of intravenous tissue plasminogen activator (IV tPA) use and primary occlusion site on the Thrombolysis in Cerebral Infarction (TICI) score. Results: Thirty-one patients were excluded for infarctions not involving the MCA territory. A total of 178 patients were included in the study. Sixty-eight patients (38.6%) had infarctions of less than one-third of the MCA territory, 50 (28.4%) had infarctions between one-third and two-thirds, and 58 (33%) had infarctions involving greater than two-thirds with 54.3% suffering infarctions of the left side. Only four patients (2.2%) required a hemicraniectomy with no statistically significant association found between TICI score and hemicraniectomy (p=0.41) or between administration of IV tPA and hemicraniectomy (p=0.36). The primary occlusion site was found to influence TICI score (p=0.045). Conclusion: A very small number of patients required hemicraniectomy after IAT as compared to previously published rates in the literature. However, several factors may prevent the patient from being an appropriate hemicraniectomy candidate in the first place and the small number of these patients in this study limits statistical analysis. The variables that determine a patient’s candidacy for decompressive hemicraniectomy remains multi-factorial.


Introduction
Acute ischemic stroke (AIS) is one of the leading causes of morbidity in the United States [1]. However, recent technical improvements in thrombectomy devices have led to higher recanalization rates [2][3][4][5]. Increased rates of recanalization are associated with higher rates of patient survival, including improved functional outcomes [6,7].
The Thrombolysis in Cerebral Infarction (TICI) grading system was designed to be the standard for reporting the results of intra-arterial thrombectomy (IAT) following AIS [8]. Recent studies have examined the TICI grading system to determine what should define "successful" reperfusion [9][10][11]. The desired outcome of treatment is complete reperfusion; however, the differences in functional outcomes and clinical relevance based on reperfusion scores is still debated [9,10,12].
Decompressive hemicraniectomy is a procedure in which a large bone flap is removed to allow the brain to expand extracranially, alleviating increased intracranial pressure, and has been shown to be a life-saving procedure in the setting of complete middle cerebral artery (MCA) infarction [13][14][15][16]. Following IAT, decompressive hemicraniectomy may be used as emergent management for malignant cerebral edema and/or hemorrhage associated with the infarction [17,18]. It has been suggested that innovations in endovascular techniques may now be decreasing the incidence of hemicraniectomy [19][20][21][22].
The primary outcome measures of this study was to determine the incidence of patients undergoing decompressive hemicraniectomy following IAT and whether TICI score affects the rates of hemicraniectomy after thrombectomy as a surrogate for stroke associated morbidity. We hypothesize that even without complete recanalization, patients who obtain any significant reperfusion are at decreased risk of requiring 1 1 1 1 1 1 1 decompressive hemicraniectomy. Additional outcomes of interest included correlation of hemicraniectomy with use of intravenous tissue plasminogen activator (IV tPA) and with site of vascular occlusion. Overall, mortality was also investigated as an outcome of interest correlated with TICI score, IV tPA use, and site of vascular occlusion.

Study population
Institutional review board approval was obtained for this study (IRB#x170424003). A retrospective chart review was completed for all consecutive patients with infarction involving the MCA vascular territory who underwent IAT between September 2014 and May 2017. The patients were then divided based on TICI score achieved following IAT. Grade 0 was considered no perfusion, grade 1 was considered penetration of thrombus with minimal perfusion, grade 2a was considered less than 2/3 of the entire vascular territory visualized, grade 2b was considered greater than 2/3 of the entire vascular territory visualized, and grade 3 was considered complete perfusion of the vascular territory. Data on age, sex, race, mortality, and comorbidities (e.g. tobacco use, atrial fibrillation, diabetes, hypertension, hyperlipidemia, and coronary artery disease) were also collected. The primary outcome was the need for decompressive hemicraniectomy. The secondary outcome of patient mortality was investigated with TICI score, IV tPA use, and site of occlusion considered as possible associated variables.

Image assessment
Stroke protocol MRI was used to assess the fraction of the MCA vascular territory involved, with diffusionweighted imaging used as the primary sequence of interest. Imaging was reviewed by a neurosurgery resident and a senior interventional neuroradialogist at the University of Alabama at Birmingham Department of Neurosurgery. The fraction of the MCA vascular territory involved was divided into three categories: less than one-third, one-third to two-thirds, and greater than two-thirds. The TICI score was self-reported by the neuro-interventionalist performing the procedure, as documented in the patient's chart.

Statistical analysis
Frequency distributions of demographics, clinical characteristics, and comorbidities were obtained. Chisquare tests and Fisher's exact tests, where appropriate, were used to determine the statistical significance of any differences according to TICI scores, IV tPA administration, and vital status when compared with both the need for hemicraniectomy and overall mortality. The variable representing fraction of MCA territory infarction was categorized as less than one-third, between one-third and two-thirds, and greater than twothirds MCA territory infarction. Logistic regression analysis was used to explore predictors of mortality using selected variables found to be statistically significant in bivariate analysis. Statistical significance was defined as p < 0.05. All statistical analysis were performed using SAS version 9.4 (SAS Institute Inc., Cary, NC).
38.6% (n=68) of patients had infarctions of less than one-third of the MCA territory, 28.4% (n=50) had infarctions between one-third and two-thirds, and 33.0% (n=58) had infarctions involving greater than twothirds of the MCA territory. The fraction of MCA territory infarcted was found to influence patient mortality (p=<0.0001). The infarct was found to involve the left hemisphere in 54.3% (n=94) of patients. There was no statistically significant difference based on the side of the infarct on mortality or TICI score (p=0.35 and 0.72, respectively) ( Table 1).   Seventy-six (42.7%) of patients underwent IV tPA therapy prior to endovascular treatment ( Table 2). Only four patients (2.2%) required a hemicraniectomy with no statistically significant association found between TICI score and hemicraniectomy (p=0.41) or between administration of IV tPA and hemicraniectomy (p=0.36). However, primary occlusion site was found to influence TICI score (p=0.045) but did not influence mortality (p=0.45) ( Table 3).     Of the patients receiving endovascular treatment, 38.2% (n=60) achieved TICI3, 33.8% (n=53) achieved TICI2b, 14.6% (n=23) achieved TICI2a, and 13.4% (n=21) achieved a TICI score of 0 or 1. TICI score was found to influence mortality (p=0.02) ( Table 4).

Discussion
In 2013, a consensus statement defined the target endovascular recanalization after AIS as TICI2b, which is defined as the restoration of more than two-thirds of the total downstream territory of the blood vessel affected [22]. Recent studies have re-examined this definition of successful reperfusion, obtaining significantly better outcomes when obtaining a TICI3 score [9][10][11]. In a retrospective analysis of 222 patients, Dargazanli et al. found a significant decrease in intracranial hemorrhage rates as well as better functional outcomes when achieving TICI3 compared to TICI2b [10]. In a retrospective analysis of 262 patients, Kleine et al. also found a significant improvement in neurologic outcomes when obtaining TICI3 recanalization [11].
Mechanical thrombectomy devices have been shown to save large amounts of penumbra when achieving either TICI2b or TICI3 [23]. Sporns et al. showed a decrease in the frequency of hemicraniectomy following an occlusion of the MCA after the introduction of stent retrievers, illustrating the effectiveness of these devices [21]. These devices have also led to better clinical outcomes and higher rates of recanalization when compared with IV thrombolysis [6].
In malignant MCA infarctions, decompressive hemicraniectomy can be a life-saving procedure [14]. The current study sought to investigate the rate of hemicraniectomy after IAT in patients with acute MCA vascular territory infarctions. Of a total of 178 patients that met inclusion criteria, only four patients subsequently underwent decompressive hemicraniectomy regardless of the size of MCA vascular territory infarcted. The percentage of patients undergoing subsequent hemicraniectomy in this study, 2.2%, is lower than recent studies obtaining results of 8.2% and 8.7% [21,24]. These studies together suggest that despite patients presenting with large ischemic strokes, early IAT may reduce the need for hemicraniectomy. Furthermore, reducing the need for hemicraniectomy inherently reduces the rates of subsequent cranioplasty, a procedure with a complication rate reported as high as 36.6% [25]. Therefore, early IAT may reduce the need for two additional procedures and the possible morbidity associated with each.
Caution has to be applied when interpreting these results, however, given the limitations in statistical analysis inherent to this study secondary to the small number of hemicraniectomy patients. In the clinical setting, some patients are not considered candidates for hemicraniectomy due to other medical comorbidities making them poor surgical candidates or due to family and patient wishes for end of life care. The presence of these patients in our study population could artificially reduce our numbers of decompressive hemicraniectomy after AIS.
When comparing TICI score achieved to site of vascular occlusion, a statistically significant correlation was found with more proximal occlusions having higher rates of TICI2b and TICI 3 recanalization. This relationship makes clinical sense as proximal occlusions are more easily treated by endovascular intervention. Although TICI score was not found to be statistically correlated with hemicraniectomy (p = 0.41), logistic regression analysis did show correlation with mortality (p = 0.033) with OR 0.64. This suggests that as TICI score improves, rate of mortality decreases. Additionally, IV tPA use decreased mortality (OR 0.37, p = 0.016). As fraction of MCA vascular territory infarcted increased, patient mortality increased as well (OR 3.34, p < 0.0001). These findings support what is seen clinically in patients with MCA ischemic stroke.

Conclusions
IAT may offer a protective advantage against the need for decompressive hemicraniectomy in patients, regardless of TICI reperfusion. The incidence of decompressive hemicraniectomy after AIS was not statistically correlated with TICI score, IV tPA use, or location of vascular occlusion although patient mortality was significantly associated with these same factors. Future case-control studies are needed to compare hemicraniectomy rates in patients that undergo IAT versus those who do not. Additional studies investigating outcomes when comparing variation in size of and timing for hemicraniectomy after IAT as well as cranioplasty are underway at our institution and may provide further guidance for standardization of

Additional Information Disclosures
Human subjects: All authors have confirmed that this study did not involve human participants or tissue. Animal subjects: All authors have confirmed that this study did not involve animal subjects or tissue.

Conflicts of interest:
In compliance with the ICMJE uniform disclosure form, all authors declare the following: Payment/services info: All authors have declared that no financial support was received from any organization for the submitted work. Financial relationships: All authors have declared that they have no financial relationships at present or within the previous three years with any organizations that might have an interest in the submitted work. Other relationships: All authors have declared that there are no other relationships or activities that could appear to have influenced the submitted work.